|
gptkbp:instanceOf
|
application-specific integrated circuit
|
|
gptkbp:abbreviation
|
gptkb:TPU
|
|
gptkbp:announced
|
2016
|
|
gptkbp:architecture
|
matrix multiplication optimized
|
|
gptkbp:availableOn
|
gptkb:Google_Cloud_Platform
|
|
gptkbp:cloudBased
|
gptkb:Cloud_TPU
|
|
gptkbp:developedBy
|
gptkb:Google
|
|
gptkbp:energyEfficiency
|
high
|
|
gptkbp:generation
|
gptkb:TPU_v4
gptkb:TPU_v1
gptkb:TPU_v2
gptkb:TPU_v3
|
|
https://www.w3.org/2000/01/rdf-schema#label
|
Tensor Processing Units
|
|
gptkbp:location
|
gptkb:Google_data_centers
|
|
gptkbp:marketedAs
|
gptkb:Google_Cloud
|
|
gptkbp:notableUser
|
gptkb:DeepMind
gptkb:Fitbit
gptkb:Google_Ads
gptkb:Google_Assistant
gptkb:Google_Maps
gptkb:Google_News
gptkb:Google_Play
gptkb:Nest
gptkb:YouTube
gptkb:Google_Cloud_AI
gptkb:Calico
gptkb:Google_Research
gptkb:Verily
gptkb:Waymo
gptkb:Google_Shopping
|
|
gptkbp:openSource
|
TensorFlow Lite for Edge TPUs
|
|
gptkbp:performanceContext
|
high throughput
|
|
gptkbp:platform
|
ASIC
|
|
gptkbp:purpose
|
accelerate machine learning workloads
|
|
gptkbp:relatedTo
|
gptkb:Field_Programmable_Gate_Arrays
Central Processing Units
Graphics Processing Units
|
|
gptkbp:supports
|
gptkb:TensorFlow
bfloat16
float32
int8
|
|
gptkbp:usedFor
|
gptkb:artificial_intelligence
deep learning
neural network training
neural network inference
|
|
gptkbp:usedIn
|
gptkb:Google_Photos
gptkb:Google_Search
gptkb:Google_Translate
gptkb:AlphaGo
|
|
gptkbp:bfsParent
|
gptkb:TPUs
|
|
gptkbp:bfsLayer
|
6
|